IAEA-SM-352/28 THE LONG TERM STORAGE OF ADVANCED GAS-COOLED REACTOR (AGR) FUEL XA9951796 P.N. STANDRING Thorp Technical Department, British Nuclear Fuels pic, , Seascale, Cumbria, United Kingdom

Abstract

The approach being taken by BNFL in managing the AGR lifetime spent fuel arisings from reactors is given. Interim storage for up to 80 years is envisaged for fuel delivered beyond the life of the Thorp reprocessing plant. Adopting a policy of using existing facilities, to comply with the principles of waste minimisation, has defined the development requirements to demonstrate that this approach can be undertaken safely and business issues can be addressed. The major safety issues are the long term integrity of both the fuel being stored and structure it is being stored in. Business related issues reflect long term interactions with the rest of the Sellafield site and storage optimisation. Examples of the developement programme in each of these areas is given.

1. INTRODUCTION

British Nuclear Fuels (BNFL) has been contracted to manage the lifetime irradiated AGR fuel arisings from British Energy reactors1. The agreement formulated is a mixture of reprocessing (covering the planned life of the Thorp reprocessing plant) and interim storage for the remainder of the fuel arisings. Interim storage is projected to be up to 80 years to comply with direct disposal acceptance criteria and projected repository availability. Eighty years represents a significant increase in storage times compared to current operational experience; of around 18 years. Confidence that AGR fuel can be stored safely for extended periods has been provided by our experience of storing AGR fuel to date and the supporting research and development programmes initiated in the late 1970's for wet storage and 1990's in the case of (SNL) dry storage project.

AGR fuel elements comprise 36 stainless steel clad fuel pins, containing uranium dioxide fuel pellets, are held together by stainless steel braces enclosed in an open ended graphite sleeve which acts as part of the neutron moderator. Normally 8 fuel elements (7 in the case of Dungeness NPP) are held together by a tie rod running through the central tube of each fuel element to make up what is referred to as an AGR stringer. After irradiation the stringer is dismantled into individual fuel elements before being wet stored in fuel skips.

The main differences between AGR long term storage as proposed as part of the SNL dry storage project and BNFL (Sellafield) taking on such a contract are:- SNL was limited by reactor site operating licences (which do not allow the transfer of fuel between sites, and therefore the need for a store at each reactor site), ARTICLE 372 effectively does not allow the removal of graphite sleeve, (and simplification of the dry store technology) as this would be viewed as waste, and existing storage facilities are limited. In comparison BNFL (Sellafield) already manages fuel from national and international reactors prior to reprocessing, operates fuel dismantling and associated waste store facilities and has three existing large pools (AGR Storage Pond, Fuel Handling Plant and Receipt & Storage) which are utilised for AGR fuel storage.

2. PROPOSED STRATEGY

Figure 1 outlines the available storage options and the main Pros and Cons of each. Based on compliance with the principles. of waste minimisation and the avoidance of high initial capital

1 British Energy reactors comprise of the former generating companies and Scottish Nuclear Limited. 2 ARTICLE 37 of the EURATOM Treaty is plant specific, it relates to the impact of operations on neighbouring member European states. Changes to ARTICLE 37 would require reapplication.

215 STANDRING expenditure the approach to be taken in the first instance is to use existing facilities and storage techniques; i.e. wet storage. This approach dictates development requirements to demonstrate that long term storage can be carried-out safely and to resolve business related issues. These can basically be divided into two main development areas:- Fuel Integrity and Pool Storage Management.

Decision For/Against

-Conforms to the principles of waste minimisation -No major initial capital investment and cost of any modifications can be phased Use of existing facilities - Existing facilities with operating licences (e.g. Fuel Handling Plant, AGR Storage Pond) (Receipt & Storage )

AGR Interim Storage - Need to demonstrate viability for extended life - Storage regimes reflect safety issues related to frequency of handling - Not stand alone units

-Built to latest building standards -Designed for maximum storage duration -Designed for storage only -Stand alone

Purpose built interim store -High initial capital investment (either dry or wet) -Does not conform to the principles of waste minimisation -Application for planning permission may not be granted -Operating licence inquiry/Public relation issues

FIG. 1. Available storage options

If the initial option becomes untenable on safety grounds the alternative of constructing a purpose built facility will be taken. The decision then will either be to go down the existing wet technology route or to develop a variation of the SNL dry store technology.

3. FUEL INTEGRITY

Spent AGR fuel cladding performs two functions; as primary containment barrier, and for mechanical handling of individual fuel pins for rod consolidation/reprocessing purposes. The final recovery and conditioning of the fuel after 80 years storage in principle is the more restrictive of the functions, based on a minimum requirement of half original wall thickness, compared to localised cladding perforations which can be resolved by encapsulation.

Figure 2 outlines the bounding cladding thickness, the margin between post irradiation and minimum clad thickness to allow for mechanical handling, and provides a very simplistic comparison with the long term wet storage of BWR fuel. Whilst wet storage of zircaloy clad fuel can be considered to be unlimited, in the absence of any other failure mechanism except general corrosion, AGR wet fuel storage is limited to a maximum predicted to be -152 years or even less (see below).

It has been well reported [1-3], that irradiated AGR fuel elements 1-5 of the original irradiated stringer are known to be susceptible to irradiation induced intergrannular stress corrosion cracking of the stainless steel fuel cladding and structural components. To inhibit this failure mechanism AGR fuel is stored in pool water dosed with sodium hydroxide to pH 11.5. Sodium hydroxide was chosen as a result of a corrosion inhibitor development programme undertaken early 1980s and has been used since 1986 for the interim storage of all AGR fuel at Sellafield. Operational experience to date indicates that fuel cladding perforation has been totally prevented.

216 IAEA-SM-352/28 The technical case for the storage of AGR fuel for up to 80 years is reliant upon the continued use of corrosion inhibitors. With changes to the front end of the fuel cycle, such as increasing fuel burnup, combined with a significant increase in proposed storage duration there is a need to revisit the original corrosion development work. One aspect of AGR fuel corrosion to be investigated as part of the development programme is the underlying corrosion rate of stainless steel in sodium hydroxide to pH 11.5.

BWR AGR

860 y.m as manufactured 388 \im as manufactured

750 ^m worst case post pile 270 (im worst case post pile

Wet storage in deionised water Wet storage in deionised water general corrosion rate general corrosion rate 0.5 i/ 7.3 x l(T'(im/year [4] [5]

430 Jim half original wail thickness 194 |im half original wall thickness assumed worst case to allow for worst case to allow for mechanical handling mechanical handling Maximum wet storage duration 4.4 xlO10 years " 152 years"

U assumes storage in deionised water and there are no other failure mechanism except general corrosion

FIG. 2. Simplistic comparison of maximum wet storage duration

During the original corrosion inhibitor development programme it was noted that there was an additional anodic (corrosion) reaction occurring in caustic solutions. Limited work was undertaken which identified that this reaction was associated with enhanced dissolution of the surface oxide layer. As this reaction was not associated with localised corrosion, and because pool storage duration prior to reprocessing was typically ~ 10 years it was not considered to constitute a threat to the integrity of either the cladding or braces and was therefore never quantified.

Given the current storage duration being proposed, such low corrosion rates could ultimately result in significant reduction in the cladding thickness. A Direct Current Potential Drop technique Field Signature Method (FSM) is currently being used to establish the general corrosion rate in sodium hydroxide; the principles of this technique are given in Appendix 1. Given a minimum resolution of 0.05% of metal thickness, a monitoring period of around a year (~ 0.19 urn limit of detection over measuring period) will be enough to establish whether general corrosion in sodium hydroxide is an issue. In the event of general corrosion in sodium hydroxide not meeting the safety criteria then alternative corrosion inhibitors will be investigated in the first instance.

In the longer term some form of condition monitoring during the long term storage of AGR fuel is required; both the condition of the fuel pins and the fuel element braces needing to be assured. Currently the integrity of dismantled AGR fuel stored in the AGR Storage Pond is monitored by means of an activity release model. The technique is retrospective and the application of non destructive techniques such as Electrochemical Noise (see Appendix 1) and FSM to give predictive information will be investigated as alternative methods.

217 STANDRING 4. POOL STORAGE MANAGEMENT

The use of existing facilities primarily raises two issues:- Firstly the need to provide confidence in the integrity of existing facilities for such extended periods of time. Secondly the fuel storage facilities were designed as buffer facilities for reprocessing activities; i.e. high throughputs and reliance on the site infrastructure. The following development programmes have been initiated to address these issues.

4.1. Structural integrity

Structural integrity of the existing facilities is crucial to a re-use policy. All ancillary plant items such as cranes, electrical wiring, roofing etc. can all be either refurbished, up-graded or replaced. Reinforced concrete is prone to the following deterioration mechanisms:

• Direct corrosion of the steel reinforcement Via cracking and spalling concrete, allowing the moisture in the atmosphere to attack the steel; • Carbonation of concrete Carbon dioxide in the atmosphere diffuses into the concrete and chemically reacts with the products of cement hydration, increasing the porosity of the structure, opening the way to steel corrosion; • Chloride ingress Chloride is the most common material which can destroy the protective passivation of he steel in concrete and is generally provided from the atmosphere; this is especially relevant in Sellafield's case due to its coastal location. The loss of passivation again open the way to steel corrosion.

To establish the current status of oxide pool facilities at Sellafield structural surveys have been undertaken specifically looking for any signs of either physical or chemical /environment effects which would impact on their long term integrity. All pools have been found to be in good condition.

The next stage in the investigative process is to extract shallow small diameter test cores from candidate facilities to determine:- characteristic cube strength, depth of carbonation, chloride ingress/penetration, and chemical/petrological examinations to establish cement contents/ waterxement ratios. The use of Fick's Law of Diffusion will then be used to give an indication of the time it will take for the carbonation front to arrive at the reinforcement.

The right choice of facility, however, is also dependent on not only its general condition, but on the evidence of how it was built. Facilities built post 1980 (Fuel Handling Plant and Receipt & Storage) benefit from improved quality assurance, standard of build e.g. seismic qualification, a change in emphasis to durability and better building techniques. The Receipt & Storage facility represents the peak of these achievements and has collected three awards including a special award from the Institution of Structural Engineers (1990) in recognition of "construction to the highest modern quality and safety standards".

Condition monitoring is part and parcel of both the site and facility operating licences. To ensure the long term integrity of the chosen facility is maintained, improvements to the current routine visual inspections are being investigated. These include non destructive techniques such as dynamic monitoring systems, for example Modal Testing Techniques, to initially establish fingerprints of the pool structures which can then be monitored, ideally remotely on line, to provide an early indication of potential system failure.

4.2. Operational

4.2.1. Current process

Irradiated AGR fuel elements are received at Sellafield into Fuel Handling Plant in 15 compartment (A2) skips which are placed into lidded storage containers. Each skip compartment

218 IAEA-SM-352/28 contains a single fuel element. After an initial period of cooling to reduce decay heat the fuel is dismantled and the resulting fuel pins are packed into stainless steel slotted canisters. Each canister contains the equivalent of three elements (108 rods) and they are placed in 20 compartment skips (Al) which are also stored in locked lidded containers. This achieves an overall four fold increase in storage capacity. The dismantled fuel is then transferred to the AGR Storage Pond until it reaches a minimum of 3 years cooled. When required, the fuel is transferred to Receipt & Storage for reprocessing.

The function of skips being housed in locked lidded containers is to prevent a criticality under the fault scenario of a three high stack being knocked over and the dismantled fuel contents spilling out. This arrangement also facilitates the stacking of containers and has the ability to control container water chemistry. All storage facilities are reliant on up stream deionised water plants as the source of pool purge water and down stream sentencing or conditioning plants for pool purge effluents.

4.2.2. Move to long-term storage

In comparison with current operations the longer term requirements would be to make the chosen facility independent of others and incur low operating costs. For example the pond purge from Fuel Handling Plant is discharged to sea via the Site Ion Exchange Plant (current availability to around 2019), and to reduce resource levels to a minimum by the incorporation of remote techniques to monitor plant status. In terms of storage layout there are significant areas for improvement as the need to be able to immediately access any particular container of fuel is no longer required. Additionally there is no longer a requirement to dismantle the AGR fuel, i.e. removal of the graphite sleeve, unless there is a net benefit through fuel consolidation activities.

Preliminary engineering studies have looked at current storage regimes and the potential for optimisation to deal with the projected business needs. For example if the fuel is stored undismantled in (A2) skip/container combinations there would be a three fold reduction in potential storage capacity. A review of criticality for undismantled fuel allows the requirement for containedsation to be relaxed to just a fixed lid on a skip. Under these conditions the skips can touch both horizontally and vertically without compromising the criticality safety case.

Given the above conditions, i.e. removal of containers and close packing of skips, the standard grid positions in one bay of Fuel handling Plant can be almost doubled. In addition stack height can be increased from three high for containerised storage to a potential four high for skip only storage; with no change to the shielding requirements. The net impact of these changes would mean that the fuel could be stored undismantled without any loss in overall capacity of the facility (in weight of uranium terms).

The question now remains whether storage capacity of the facility could be further increased if Burnup Credit was approved by applying the same arrangements to dismantled fuel.

5. SUMMARY

The initial approach and fall back positions being taken by BNFL in managing the AGR lifetime spent fuel arisings from British Energy reactors has been out-lined. Interim storage for up to 80 years is envisaged for fuel delivered beyond the life of the Thorp reprocessing plant. Adopting a policy of using existing facilities has defined the development requirements to demonstrate that this approach can be undertaken safely and business issues can be addressed. The major safety issues are the long term integrity of both the fuel being stored and structure in which it is being stored. Business related issues are reflected in the need to address long term interactions with the rest of site and storage optimisation.

219 STANDRING REFERENCES

[1] INTERNATIONAL ATOMIC ENERGY AGENCY, Extended Storage of spent fuel, IAEA- TECDOC-673, Vienna (1992). [2] INTERNATIONAL ATOMIC ENERGY AGENCY, Further analysis of extended storage of spent fuel, IAEA-TECDOC-944, Vienna (1997). [3] HANDS B.J., "UK (BNFL) practices on wet fuel storage at Sellafield", Storage in Nuclear Fuel Cycle, C512/039/96 BNES, London (1996). [4] AMERICAN NUCLEAR SOCIETY, Long term storage of spent fuel, ANS 206, (1986). [5] WARNER B.F., The storage in water of irradiated oxide fuel elements, Technical data provided to the Windscale Inquiry, AEA, Windscale (1977).

APPENDIX 1

The principles of Field Signature Method (CorrOcean™) experimental technique

The Field Signature Method (FSM) is based upon feeding a direct current through the object to be monitored and measuring the resultant electric field via an array of electrical contacts on the external surface of the object. Changes to the magnitude of these electric fields result from any metal loss due to corrosion. The voltage As (pin pair on structure being monitored) and Bs (reference pair) are measured with the structure in its initial condition. This is the "signature" reading. Subsequently voltages A* and B; are measured for the same pairs of electrodes and a Fingerprint coefficient is calculated using the equation:-

FCM= {BS xA;- 1} x 1000

{As Bj } (parts per thousand) where

Fc^ = Fingerprint coefficient for pair A at time ,

As = Voltage across pair A at start, Bs = Voltage across reference pair B at start, Ai = Voltage across p[air A at time i, Bj = Voltage across reference pair B at time i. By using voltage ratios any variation in the excitation current is automatically compensated. The potential drop between electrode pairs is typically 100-200 |iV with a minimum of 50 |j.V being recommended by CorrOcean. The equipment has a measuring resolution of 4.77 nV. A resolution of 0.05% of the wall thickness is claimed for general corrosion. With filtering to remove noise a resolution of 0.01% is expected to be achievable.

National Nuclear Corporation (NNC) Electrochemical Noise Probe

NNC Electrochemical Noise Probes were used as part of fuel integrity development work to support the SNL dry store project. The technique basically measures fluctuations in the corrosion current and potential (Electrochemical Noise) on a suitable corrosion probe as a result of changes in environmental conditions. The technique is reliant on producing probes that are representative of the material to be long term stored.

The potential application of this technique would be in the form of an on-line monitor to provide a quick response to loss of pond water chemistry causing corrosion to initiate. It is not expect that this technique would measure the extent of attack, but would compliment the FSM technique.

220